Will develop a dynamic model of response to HER family inhibitors in tumors in which HER2 is amplified that encompasses both fast, phosphorylation-based events (on the order of a few minutes) and slower transcriptional and epigenomic processes (on the order of a few days). This model will eventually enable comparative assessment ofthe relative importance of mechanisms of response and resistance and guide development of combinatorial therapeutic strategies to counter resistance. This project is motivated by observations that responses to trastuzumab and lapatinib are not uniform between patients and are frequently not durable. Work in this CCSB project and the general scientific community suggests several mechanisms that may confer resistance including: (a) activating downstream mutations in the PI3K pathway, (b) microenvironment mediated activation of interacting networks, (c) PI3K mediated changes in HERS expression and signaling and (d) transcriptional feedback regulation from response related network elements. An initial dynamic version ofthe model will be developed in collaboration with the MIT CCSB (see letter of collaboration from Dr. Lauffenburger). The model will differ from existing work in three important ways: it will exploit a mathematical separation of time scales for fast and slow dynamics, incorporate underlying genetic aberrations, and include parallel signaling from the microenvironment. Analysis ofthis initial model will be used to help understand the roles of cooperating genetic aberrations, transcriptional and translational regulation, vesicle control and microenvironment in fast and slow dynamic processes. Subsequent versions of the model will build on experimental measurements of temporal biological and molecular responses of HER2+ breast cancer cell lines to HER2 family signaling network inhibitors administered alone and in combination as well as information from Projects 1, 2 and 4 and from the Stanford CCSB's MYC modeling efforts (see letter of collaboration from Dr. Plevritis). A combination of Bayesian network analysis and dynamic modeling will be used to model the unexplored effects of epigenomic modulation of transcription on HER2/3 signaling.